3.1.1 How to combine experts’ assessments?

Course subject(s) Module 3. Performance-based weights and the Decision Maker

Recall that during the first module, it was mentioned that a desiderata of structured expert judgement is to use the assessments of multiple experts.

The question that was raised then was: how can one combine experts’ assessments?

Back to the Dutch eating habits example. Recall that, in the example, a Dutch supermarket was interested in the eating habits among Dutch adults.

Suppose the purpose of using structured expert judgment are the prediction of future eating habits among Dutch adults, say in 2025. And suppose that the interest is in the following questions of interest

6) What percentage of Dutch adults will eat fruit on a daily basis by 2025?
7) How many liters of milk will be consumed on a yearly basis by the average Dutch adult in 2025?

Note that, as beforehand, the data are purely fictional. Also note that for the questions of interest, there is no realization yet!


Question Realization Expert 1

5%  50%  95%

Expert 2

5%  50% 95%

Expert 3

5%  50%  95%

6 61   68   75 45    55    70 43    52    63
7 67   71   80 50    61    68 55    65    73

For each question of interest we have 3 pairs of assessments. How can we combine them?

During the previous module, you learned how to evaluate experts’ assessments using two scores; the calibration and the information score. At this point in the course we will show how the two scores can also be used in combining. During this module, you will learn how to combine experts’ assessments.

Consequently, we will be able to provide answers to the questions of interest.

Creative Commons License
Decision Making Under Uncertainty: Introduction to Structured Expert Judgment by TU Delft OpenCourseWare is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Based on a work at https://online-learning.tudelft.nl/courses/decision-making-under-uncertainty-introduction-to-structured-expert-judgment//.
Back to top